Research on the Influence of Memory Size on the Performance of Multi-Tasking Virtual Machine Live Migration

• Main Authors: Dong-Ye Xie, 謝東曄
• Other Authors: Cho-Chin Lin
• Format: Others
• Language: zh-TW
• Published: 2018
• Online Access: http://ndltd.ncl.edu.tw/handle/9z99c8

碩士 === 國立宜蘭大學 === 電子工程學系碩士班 === 106 === Virtual machine live migration brings many benefits to cloud service providers, such as load balancing and more efficient use energy. The two main performance indicators for live migration are the total migration time and service downtime. The shorter the time is, the higher the performance will be. The service downtime is important from the user’s view point since a long service down time may result in a severe impact on the service level agreement cosigned by users and service providers. The impact of the problem can be minimized by shrinking memory space available to the tasks running on the virtual machine. In additional, multitasking can also be employed by virtual machines to reduce the adverse effect of resource underutilization. In this paper, we study the performance of multitasking virtual machines which migrate under various sizes of available memory. The performance metric is the time taken by the multitasking virtual machine to complete the concurrent tasks which start at a host and finish at another. Our experimental results show that the performance can be improved by appropriately packing tasks of various types for multitasking under reasonable shrunk memory size.